528 PROFESSOR KNOTT ON THE STRAINS PRODUCED IN 



were cut, turned, and bored by Messrs Aitken & Allan, Edinburgh, who also supplied 

 the iron and steel tubes. 



§ 2. The Tubes. — It was hoped that a careful study of the behaviour of these tubes 

 in various magnetic fields would throw some light on the character of the strains 

 accompanying magnetisation in iron and nickel. In the preliminary results for steel 

 and nickel, however, no very evident relation could be detected among the tubes of 

 different bore, other conditions being the same. In the case of iron, a gradation in 

 the law governing the changes of volume was no doubt perceptible as we passed in 

 succession from tube to tube. But, in general, it was evident that the different tubes 

 of any one metal were not identical as regards homogeneity and other physical qualities, 

 notwithstanding the fact that all had been cut from the same bar. There seemed to be, 

 accordingly, only one certain method of studying the influence of bore upon the strain 

 produced in any given field — and that was, to use throughout exactly the same tube. 

 Thus for each metal the tube of narrowest bore was carefully studied, the changes of 

 internal volume at different fields being measured in a way hereafter to be described. 

 This tube was then bored out to a larger diameter of bore, and in this new state 

 similarly treated in the magnetic field. The necessary measurements made, the tube was 

 again bored out to a still larger diameter of bore, and once more subjected to the same 

 treatment ; and so on in succession, until the widest bore that could conveniently be 

 used was reached. In this way the absolute identity of material of the walls of the 

 different tubes was secured. From each tube of narrowest bore, six other tubes were 

 formed by successive borings. These are distinguished in the present paper by the 

 numbers I. to VII. # The other tubes of permanent bore, three in the case of each, will 

 be distinguished by the Indian numerals. The particular number chosen for any tube 

 has the same name as the Eoman numeral for the tube which is nearest to it in dimen- 

 sions. Thus the iron and steel tubes have Nos. 3, 5, 7, and the nickel tubes, 1, 4, 7. 



§ 3. Fixing op Capillary to Tube. — Each tube was bored out to a short distance 

 from the bottom. In addition to the uniform bore throughout, a wider but shallow 

 part was removed at the top, and a screw cut in it. Into this screw a cap of iron, steel, 

 nickel, or brass could be fitted, and through a central hole in the cap passed the capillary 

 tube, which afforded the means of measuring the change of volume. Much time and 

 thought were spent, and many unsatisfactory experiments were made, before a con- 

 venient form of cap was devised. The enumeration of a few of the requirements 

 to facilitate operations, and a few of the most troublesome sources of error which 

 had to be guarded against, together with a brief statement of how these requirements 

 were met and difficulties overcome, will suffice as a description of the final form of cap 

 adopted. 



In the first place, it was essential to have a ready means of adjusting the capillary to 

 each tube in turn. In the preliminary experiments with iron the capillary tube was 



* These must not be confused with the numbers in the previous notes dealing with the permanent tubea of 

 different bore. 



